Protein homologues often have common functions, with variation evolving via amino acid substitutions at specificity determinants (SDs). These residues may be located in binding sites or other "long-range" regions of the protein. Predicting SDs is a current target of bioinformatics sequence analyses. Our long- term goal is to experimentally identify long-range SDs in the Lacl/GaIR family;the proposed work focuses on SDs in the linker that connects the DMA-binding domain to the regulatory domain. Hypotheses are derived from 4 different prediction strategies, which are only in partial agreement with each other. The common function of the Lacl/GaIR proteins - transcription control - allows "moderate through-put" assays of function so that all predictions may be compared. We will test hypotheses in multiple homologues to address the question: Do homologues utilize all SDs available to the common fold (a frequent assumption of prediction algorithms) or do different functions require only a subset of potential SDs? Further, the aspect of function affected by changing an SD cannot yet be reliably predicted, nor is it clear whether a functional change for one SD is the same in all homologues. Our experiments will monitor different aspects of Lacl/GaIR function, including DMA specificity, DNA affinity, and allosteric response to binding regulatory effector molecules. Proposed experiments utilize chimeras comprising the Lacl DMA-binding domain and regulatory domains from E. coli paralogues. Linkers come from Lacl or paralogues. Since each naturally-occurring Lacl/GaIR protein recognizes a different DNA ligand, the common DNA-binding domain allows us to more easily parse functional contributions from binding site and long-range SDs. By definition, making an amino acid substitu- tion at an SD will change function. We will use in vivo repression/response to effector and in vitro thermody- namic measurements of affinity/allosteric response to characterize the chimeras and potential SD variants. Specificity for alternative DNA ligands will be determined. Experiments are designed to answer the following questions: Aim 1: Can one linker facilitate allosteric communication with a variety of regulatory domains? [or do altered linker SD interactions abolish this function?] Aim 2: What are the functional contributions from specific positions in the linker? Aim 3: Can knowledge of SDs [in one linker] be used to transplant lacO1 DNA-binding to other linkers? Results will: (1) Identify the locations of SDs and determine if they make similar functional contributions to several homologues;(2) Yield a list of empirical rules for creating novel Lacl/GaIR proteins with biotechnological utility via domain recombination;and (3) Test the current prediction algorithms and generate a new sequence/function database for improving predictions. This work will facilitate expanded use of data generated by the Human Genome Project.